Software-Defined Vehicles: How Over-the-Air Updates Restructured Profitability

# Software-Defined Vehicles: How Over-the-Air Updates Restructured Profitability

"Software-defined vehicle" is one of those industry phrases that has been used so broadly it's lost precision. Every car has software. The question is whether the software architecture enables fundamentally different business models and vehicle improvement trajectories than traditional automotive electronics.

## What SDV Actually Means

Traditional vehicle electronics are *distributed and domain-segregated*: dozens of separate Electronic Control Units (ECUs) each handling a specific function — one for the engine, one for the ABS, one for the climate control, each with proprietary communication protocols. Software updates required physical connection, specialist tools, and in many cases dealership visits. Adding a new feature often required hardware changes.

*Zone architecture with centralized compute* changes this: instead of dozens of specialized ECUs, a smaller number of high-performance computers (often called "zonal controllers" or "central domain controllers") run multiple vehicle functions in software. Tesla's Full Self-Driving computer runs vision processing, decision-making, and motion planning as software on shared hardware. Updates to that software improve capability without touching the hardware.

The business model implication is significant: if most vehicle features are software, they can be upgraded, unlocked, or sold after vehicle purchase. Tesla's FSD (Full Self-Driving) software is sold as a subscription or purchase add-on. The physical hardware is already in the car — you're paying for software access.

## How Tesla's OTA Capability Changed the Depreciation Curve

Traditional ICE vehicles depreciate primarily on physical wear. A 5-year-old BMW 3 Series has the same feature set it shipped with — depreciation reflects accumulated mileage and component aging.

A 5-year-old Tesla Model 3 may have better autonomous driving capability, better charging speed management, better range efficiency, and better infotainment than it did when new. OTA updates have added features (V2G capabilities, improved Autopilot algorithms, new parking features) and improved existing ones.

This changes the residual value calculation. The data supports this: Tesla vehicles have held residual value better than most comparable luxury vehicles in the years since widespread OTA capability. Whether that continues as competition intensifies is unclear, but the basic mechanism is real.

The depreciation comparison is unfair to traditional OEMs in one important respect: they've been constrained by hardware architectures designed before OTA was expected. A 2019 Mercedes E-Class has dozens of separate ECUs that can't be easily updated over the air — not because Mercedes didn't want that capability, but because the car wasn't designed for it. Legacy OEMs are rebuilding their architectures from the ground up (Volkswagen's VW.OS, GM's Ultium platform software, Stellantis EV platforms), which takes 5-10 years to fully deploy.

## Why Legacy OEMs Struggle to Replicate FSD as a Revenue Stream

Ford and GM have announced various software subscription services, and most have struggled or been cancelled. The challenges are structural:

**Dealer network conflict**: Legacy OEM revenue flows through franchised dealers who earn margin on service. OTA updates that prevent service visits reduce dealer service revenue. Dealer networks have historically lobbied against features that eliminate service touchpoints.

**Hardware fragmentation**: A fleet of 50,000 vehicles sold over 5 years may have 20 different hardware variants if the underlying ECUs were sourced from different suppliers across model years. An OTA update strategy requires software compatibility across that fragmentation.

**Liability architecture**: When an OTA update changes vehicle behavior, who bears liability for incidents? Traditional automotive liability flows through physical design processes with defined testing requirements. OTA-updated software changes the vehicle post-certification, which creates a legal framework mismatch.

## The Honest Tension

More software means more capability — and more attack surface. Automotive cybersecurity researchers have demonstrated remote compromise of vehicle systems through multiple vectors. The increasing connectivity and central compute architecture that enables OTA updates also creates entry points that didn't exist in distributed ECU designs.

The privacy implications are significant: a connected vehicle generates an enormous amount of data about driving behavior, location, and occupant activity. Tesla's data collection has been scrutinized by regulators in multiple countries. The data is valuable for product improvement and training autonomous systems; it's also sensitive personal information that raises legitimate privacy concerns.

The SDV transition is happening regardless — the capability gains are too significant for OEMs to ignore and customers are demanding the features. Managing the security and privacy implications is an ongoing problem, not a solved one.

Software-Defined Vehicles: How Over-the-Air Updates Restructured Profitability

// COMMENTS

ON THIS PAGE